Cisco Nexus 7000 Series Configuration Manual page 88

Graceful Restart PIM and Multicast Traffic Flow
Figure 16: Operation of Graceful Restart PIM during an RP Switchover
The mechanics of the support for Graceful Restart PIM feature are as follows:
• In steady state, PIM neighbors exchange periodic PIM hello messages.
• An active RP receives PIM joins periodically to refresh multicast-route (mroute) states.
• When an active RP fails, the standby RP takes over to become the new active RP.
• The new active RP then modifies the Generation ID (GenID) value and sends the new GenID in PIM
• Adjacent PIM neighbors that receive PIM hello messages on an interface with a new GenID send graceful
• Those mroute states are then immediately reestablished on the newly active RP.
Graceful Restart PIM and Multicast Traffic Flow
Multicast traffic flow on PIM neighbors is not affected if the multicast traffic detects support for Graceful
Restart PIM or PIM hello message from a node with the failing RP within the default PIM hello hold-time
interval. Multicast traffic flow on a failing RP is not affected if it is Non-Stop Forwarding (NSF) capable.
Caution The default PIM hello hold-time interval is 3.5 times the PIM hello period. Multicast High Availability (HA)
operations may not function as per design if you configure PIM hello interval with a value lower than the
default value of 30 seconds.
Cisco Nexus 7000 Series NX-OS Multicast Routing Configuration Guide
74
hello messages to adjacent PIM neighbors.
restart PIM for all (*, G) and (S, G) mroutes that use that interfaces as an RPF interface.
Configuring PIM and PIM6